The Non-Catalytic Polymerization of Ethylene Dissolved in Naphthalene

Home , Naphthalene

- m

THE NON-CATALYTIC POLYMERIZATION OF ETHYLENE DISSOLVED IN NAPHTHALENE

Clyde Kempton Smith A.B., Stanford University 1933

and

Ben T. Woodruff B.S., University of Chicago 1932

Submitted in Partial Fulfillment of the Requirements

for the degree of

MASTER OF SCIENCE

from the

Massachusetts Institute of Technology 1935

Signatures of Authors *00 0 ~ S.

Department of Chemical Engineering, July 10, 193-5 Professor in Charge of Research

Chairman of Departmental Committee on Graduate Students .... .--. .. / /'j -d f I

r. Massachusetts Inst. of Technology Cambridge, Mass. July 10, 1935

Professor W. G. Whitman Department of Chemical Engineering Massachusetts Institute of Technology Cambridge, Massachusetts

Dear Professor Whitman: In accordance with the regulations of the Department of Chemical Engineering governing thesis work, we herewith submit a thesis entitled, "The Non-

Catalytic Polymerization of Ethylene Dissolved in

Naphthalenem, in partial fulfillment of the requirements for the Master of Science Degree. Respectfully submitted,

lyde Kempton Smith

Ben T. Woodruff

203844 Acknowledgement

The authors wish to express their gratitude to Professor Hottel for his helpful advice and counsel on the investigation of this problem. They also wish to thank Mr. F. R. Russell for his willing direction and assistance in the work. TABLE OF CONTENTS

The Problem ...... 1

Summary ...... 2

0 0 0 Introduction ...... 4 Previous Work by Other Investigators 5

Summary of Procedure . . 0 0 0 ...... 0 8

Discussion of Results . . . . .0 * 0 .. 0 12

Conclusions ...... 0 0 0 * ...... 16

Recommendations ...... 0 .. 0 . 0 0 0 18 Appendix

Description of Apparatus ...... 0 0 .. 20

Procedure . * 0 0 0.. *.. 0...... 0 24

a. Operating Procedure .. . . 0 . .. 0 0 24

b. Analytical Procedure . 0...... 26

c. Calibrations . .. 0 .. 0 ... .. 0 31

Sample Calculations . 0 0 ... 0 ... 0 ... 34 Experimental Data

Literature Cited .. . .. 0 0 0 0 0 0 0 0 . 0 37 1

The purpose of this investigation was to study the polymerization of ethylene in the presence of an inert, non-catalytic hydrocarbon solvent. The solvent was to exist in a liquid phase as well as a gas phase under the conditions of reaction. Calculations were to be made to determine a constant for the rate of reaction, and some study was to be made of the products of polymerization. The problem is part of a broad program of study of the mechanism of polymerization of hydrocarbons. 2

In this investigation the primary concern was the determination of the reaction rate constant for the polymerization of ethylene dissolved in naphthalene with both liquid and gas phases present, although as a corollary the change of the constant with temperature and the nature of the products formed were also investigated. The method of finding the reaction rate constant consisted in making runs for varying lengths of time at a constant temperature of 7800 F. An indication of the change of the constant with temperature was obtained by varying the temperature while holding the time of the runs constant. By analytical means it was possible to determine the extent to which the polymerization had taken place, as well as to find in a general way the nature of the products formed. The fact that two phases were present rendered it impractical to express the ethylene concentration during the run. Another complication was the exceeding complexity of the reactions taking place in the later stages of the runs. An expression for the reaction rate constant similar in form to the second order type of reaction was obtained by plotting data on percent of original ethylene left against time of reaction. The original concentration of ethylene was about 6 mol per cent, (according to fugacity calculations). Data on the nature of the

products formed indicate that hexylene is the principal gaseous olefine formed. They indicate too that certain amounts of parraffinic hydrocarbons are formed during the reaction. 4

Introduction

Interest in polymerization of olefinic hydrocarbons has increased markedly in the past few years, particularly in the petroleum industry. It appears in cracking as troublesome coke and gum formation. On the other hand it is a very useful tool in the production of synthetic resins, solvents, and high grade motor fuels.

Especially in connection with the latter two groups of com- pounds investigators are doing extensive work.

In extending the scope of the work of polymerization of ethylene begun by Mr. Russell, it was planned to investigate the problem under conditions such that both liquid and gas phases were present. Only a very limited amount of work is reported in the literature on the non- catalytic polymerization under these conditions. Primary interest was in finding the conditions under which such polymerization would occur, in determining the reaction rate constant for the polymerization, and in gaining an indication of the change of the constant with temperature. With the limited time available it was decided to investigate the problem by arbitrarily fixing two of the four primary variables, namely, the extent of the liquid phase and the initial pressure of ethylene. Runs were made at constant temperature for varying lengths of time and for varying temperature with constant lengths of time. 5

Previous Work by Other Investigators

Although much work has been done on the gas

phase -a polymerization of ethylene, with and without

the use of catalysts, the problem of polymerization with a liquid hydrocarbon phase present has received but a limited study. A review of the literature for several years back yielded only two patents of somewhat contra- dictory nature. 2 The first one, which was issued to A. S.Ramage in 1925, states that ethylene compressed into kerosene at

00 C. and 80 lbs. per square inch pressure showed about seventy-five per cent conversion to a product which was mainly butylene. (The time of the reaction was indeter- minate from a description of the process.) He also states that higher temperatures and pressures can be used, and that the only requirement is that the ethylene dissolves in the liquid phase. Considering all the data on polymerization which requires an elevated temperature, the other patent, which was issued to H. D. Elkingtonl in 1930, seems more consistent with the facts. He states that 95 grams of ethylene heated in an autoclave with 300 grams of paraffin oil, at 45 atmospheres and 4200 C. yielded 50 grams of a light naphtha. His claims state that pure ethylene is not necessary, and that it is thus possible to use gaseous 6

cracking products oontaining ethylene, to make useful polymers.

Some of the pertinent facts concerning gas-

phase polymerization would be expected to apply to the

liquid-phase work, at least to a certain extent; they

will be discussed briefly. Ipatiev8 heated ethylene in

an iron tube under high pressure, and found that polymer-

ization began at 3250 C. and was quite rapid at 380 - 4000C. Hague and Wheeler, found that ethylene polymerized readily

between 400 and 7000 C., yielding butylene. They used a

silica bulb and heated it for three hours. According to

Sachanen and Tilicheyev12 the polymerization of olefines

to form higher olefines is thermodynamically possible up

to 5004 C. This is brought out graphically by Francis and Kleinschmidt 5 who plotted the changes in free energy occur-

ing versus the absolute temperature. The change in slope

of the curve shows a tendency for ethylene to polymerize

below 4250 C. and to crack at higher temperatures. Isomer-

ization to cycloparaffins is possible below 4250 C.

Other facts, some of which were coroborated by

this investigation, are mentioned briefly below. Pressure

seems to have only a small effect during the first stages

of the reaction. Its main purpose is to keep the ethylene

L. 7

in the liquid phase. The effects of temperature and time are similar; a short time at a high temperature shows about the same results as a long time at a lower temperature. Polymerization is generally bi-molecular, or a reaction of the second order, in contrast to cracking which is a first order process. Many chain reactions may follow initial polymerization. 8

Summary-of Procedure

The problem resolved itself into one of making various runs under pressure with a solvent present. Naph- thalene was used as the solvent because of its aromatic, non-cracking character, relative inactivity, and high critical temperature. There were four primary variables in the experiments; they may be taken as: (1) the extent of the liquid phase; (2) the concentration of ethylene in the liquid phase; (3) temperature; and (4) time of reaction.

It was obviously necessary to limit the scope of the work, and some of the variables had to be eliminated. For this reason all of the runs were made with the same liquid-gas ratio, and at the same initial concentration of ethylene.

For comparison some runs were made with only the gas phase present. The required amount of naphthalene was estimated from fugacity data to give a liquid phase of about one ninth of the reactor volume. Previous work in the literature reported a negligible effect of pressure on the reaction, and since this determines the concentration of ethylene, the amount of ethylene originally put into the system was arbitrarily chosen and kept constant. Runs at a constant temperature were made for times varying from fifteen minutes to three hours. These two variables, temperature and time, 9

were the ones mainly investigated, although a run was made to determine if the iron, of which the system mainly composed, had any catalytic effect. This was accomplished by putting steel wool in the reactor to obtain a large iron surface. The extent to which the reaction had proceeded was determined by measuring the amount of gas left after the materials had been allowed to react, and analyzing it for ethylene. The amount of ethylene disappearance indicated the extent of polymerization. Supplementing the above work, the nature of the products formed was sought as an indication of the mechanism of the reaction. An analysis of the reaction gases revealed the amounts of higher olefins, and of inert paraffinic hydrocarbons and hydrogen formed.

Gas density data were obtained both on the reaction gas, and upon the residual gas after removal of all olefins. From these results it was possible to calculate the average molecular weights of the higher olefins and of the residual gas.

Further data as to the nature of the products was obtained by a quantitative analysis of the residual gas for carbon and hydrogen by combustion.

As previously mentioned, the attack of the problem necessitated the use of high pressures. In addition to this, the high temperatures used made it necessary to employ materials which would hold their strength under the severe conditions of reaction. A high pressure reactor was available which was suitable. It was heated by a winding of resistance wire. The heads of the reactor were removed after every run to take out the mixture of naphthalene and liquid polymer, and to add fresh naphthalene. The measure of the amount of ethylene originally added to the system was made with a low range pressure gage. During all the runs where a liquid phase was present, the reactor was shaken to maintain equilibrium conditions. The procedure in making a run was to add naphthalene to the reactor, mount the latter, evacuate the system, and put the ethylene in from a cylinder under pressure.

Current was then put on the heating element of the reactor.

Initial heating was at a high rate to minimize the correction for reaction during the heating period. Temperatures were controlled by means of two thermocouples, the first one a chromel-alumel junction on the outside wall of the reactor which was used to indicate heating rate, and the second one a copper-constantan couple which extended to the center of the reactor and measured the actual operating temperature. Both pressure gages and thermocouples were calibrated against appropriate standards. Readings of temperature and pressure were made at frequent intervals during the runs. 11

Upon completion of the time previously decided upon for the length of the run, heating was discontinued, the lagging with which the reactor was covered was taken off, and an electric fan was turned on the reactor to cool it rapidly. When the conditions indicated that the naphthalene had solidified, the gas was vented off and measured. A special type of gas collection apparatus was used which allowed the measurement of the gas under atmospheric pressure. It was collected over saturated brine which was also saturated with ethylene.

A Williams gas analysis apparatus was used for the gas analysis. The first pipette contained 87% sulfuric acid to remove higher olefins. The second pipette contained bromine water which removed ethylene. The residue, after being passed through caustic solution, was assumed to be a mixture of paraffinic gases and hydrogen.

Gas-density data was obtained by weighing the gases in calibrated gas-density flasks. In determining the density of the gases inert to bromine water, these latter were separated from the olefinic gases by bubbling through towers containing bromine and caustic solutions. 9N1MMI. 1

Discussion of Results

The principal information to be drawn from this

work is in regard to the rate of reaction and the effect

of naphthalene on polymerization. The data from the con-

stant temperature runs were sufficient to allow calculation

of the reaction rate constant. The fact that two phases

were present made it difficult to express this data in

terms of concentration of ethylene. Therefore, the per

cent of the original ethylene left after a given time was

used as a basis for calculation. Another complication was

encountered in the complexity of the reactions taking place

after initial polymerization. The reaction rate constant

was figured on the assumption that during the early part

of the run the reaction was of the sinple second order type.

This gives an integration of the form:

1 = K 9 (E 2 E ) where K is the reaction rate constant, E is percent

ethylene left, and 9 is time of reaction in minutes. The

initial slope of the curve of the reciprocal of the per cent

ethylene left versus time indicated the constant. (Figs.l, 2.)

This gave a value of .00064. It was hoped to get enough

data to show how K changes with temperature, but the few runs

at constant time and varying temperature did not allow this. 13

The time used in all of these runs was so long as to be

in the portion of the curve where the assumption of a

second order reaction did not apply.

As to the effect of the naphthalene, the liquid

phase does not seem to be as important as the mere presence

of naphthalene itself. It does not seem to react with the

ethylene, because it distilled off in a fairly pure state

when the distillations were made on the naphthalene con-

taining the polymers. The increased polymerization due to

the presence of even naphthalene vapor is seen in comparing

the ethylene disappearance in runs 3 and 4. (See Calculated

Data.) From the way in which the pressure varied during the

run it appears that in pure ethylene runs an induction

period of about half an hour is required before polymeriza-

tion occurs after which it takes place fairly rapidly. (Fig.4) The converse seems to hold with the naphthalene for most of

the polymerization occurs during the first half hour. (Fig. 5)

Certain conclusions can be drawn in regard to the

nature of the products formed in the runs in which time is

a variable. The weight per cent of ethylene which polymer-

ized to heavy polymers in the naphthalene followed a curve

just the inverse of the per cent ethylene remaining. (Fig. 1) The amount of inert hydrocarbons, which were found to be a

mixture of paraffins and hydrogen, increased uniformly until

L 14

about two hours had passed and then slowly decreased.

The gaseous olefines formed were at a maximum at forty

minutes, and then decreased slowly as more of them went

to heavy polymers. The fact that the molecular weights

of the gaseous polymers were all between 65 and 85 except

in one case indicates that they were probably mixtures of

butylenes and hexylenes. The last conclusion to be drawn

is in regard to the apparent inhibitive effect of iron

surface. The data on the steel wool run (See Fig. 1 and

Calculated Data.) seems conclusive enough to indicate that a large surface of iron definitely inhibits the polymerization of ethylene, at least under the conditions of this investigation.

With regard to the reproducibility of the data, check runs were made at one half hour and constant tempera- (Runs & A&//) tureAgiving results which agreed within five per cent. Granting that the runs were all carried out with AAA amount of care as these two, this seems to be the limit of the accuracy of the data. The temperature measurements were c to better than _ degrees Fahrenheit. The assumption made that no ethylene remained in the solidified naphthalene after the run, was verified by making a check on the reactor volume. The system was filled with ethylene under operating pressure, the naphthalene was melted, and ffmmmmmmmmmlm W" r 15

and then the gas was vented off. This checked the other

calibrations within 2 per cent.

L CALCULATED DATA

Run No. 1 2 3 4 No No 23 gms. 100 gms. Conditions naphth. naphth. naphth. naphth.

Effective Time 106 min. 85.6 142 123

Effective Temperature 7500 F. 7670 F. 7500 F. 7500 F. Gas In (cu.ft.) .180 cu.ft. .262 .258 .216

Gas Out (cu.ft.) .1441 .1462 .0999 .075

.1788 .2601 .2563 .2142 C2 H4 In (cu.ft.) Analysis of Gas Out: % C2 H4 Out 83.2 75.45 63.7 77.75 % Gaseous Olefins 7.3 14.03 13.25 9.02 % Inert Hydrocarbons 9.5 10.52 23.05 13.23

C2 H4 Out (cu.ft.) .1198 .1103 .0636 .0584 % Ethylene Disappear. 33.0 57.6 70.5 72.8 Mol. Weight of Gas Out 36 33.5 29.2

Mol. Weight of Gaseous 83.5 67.9 40.0 Olefins

Mol. Weight of Inert 30.3 29.2 28.8 gases

Wt. % C2 H4 Left 67 42.4 29.5 27.2 Wt. % as Gaseous Olefins 23.32 10.07 4.5 Wt. % as Inerts 6.38 6.73 4.7 Wt. % as Liquid Polymers 28.0 53.7 63.7 Remarks No gas Small densities losses obtained of gas CALCULATED DATA

Run No. 5 7 A 8 9 100 gms. 100 gms. 100 gms. 100 gms. Conditions naphth. naphth. naphth. naphth.

Effective Time 131 min. 138 3806 68.8

Effective Temperature 8190 F. 7800 F. 7800 7800

Gas In (cu.ft.) .216 cu.ft..208 .2092 .2083

Gas Out (cu.ft.) .0428 .0655 .0853 .0887 .2069 .2078 .207 C2 H4 In (cu.ft.) .2144 Analysis of Gas Out: 40.0 51.0 75.55 73.6

% C2 H4 Out 9.34 Gaseous Olefins 5.0 10.66 12.02 Inert Hydrocarbons 55.0 38.10 12.43 17.06 .0161 .0335 .0644 .0654 C2H4 Out (cu.ft.) % Ethylene Disappearance 92.4 83.8 69.0 68.4 Mol. Weight of Gas Out 31.56 32.88 33.05 33.45

Mol. Weight of Gaseous 64.6 76.8 67.25 76.9 Olefins

Mol. Weight of Inert 31.1 27.6 30.5 32.95 Gases

Wt. % C2H4 Left 7.6 16.2 31.0 3106

Wt. % as Gaseous Olefins 2.2 9.3 11.83 11.0 Wt. % as Inerts 11.63 11.9 5.56 8.61

5 6 A5 Wt. % as Liquid Polymers 78 ._ . 51., 648761 48.79 Remarks Traps H/C analysis installed. made Small gas Ratio=3.9:1 loss I CALCULATED DATA

Run No. 10 11 12 13 100 gms. 100 gms. Naphth. 100 gms. Conditions naphth. naphth. & steel wool naphth.

Effective Time 15.4 min. 37.5 33.7 205

Effective Temperature 780 0 F. 7800 F. 780OF. 7800 F.

Gas In (cu.ft.) .2158cu. f t.2176 .2113 .2087

Gas Out (cu.ft.) .1318 .085 .1308 .0573

C2 H4 In (cu. ft. .2143 .2162 .2097 .2073 Analysis of Gas Oat: 90.57 74.7 88.34 55.0 % C2H2 4 Out % Gaseous Olefins 6.12 12.5 7.44 10.72 % Inert Hydrocarbons 3.31 12.8 4.22 34.28

.1192 .0645 .1038 .0315 C2 H4 Out (cu.ft.) % Ethylene Disappearance 44.3 70.6 45.0 85.0 Mol. Weight of Gas Out 3004 33.5 30.0 35.2

Mol. Weight of Gaseous (65.8) 70.3 (54.0) 90.4 Olefins

Mol. Weight of Inert 29.4 29.4 Gases

Wt. % C2 H4 Left 55.7 29.4 55.0 15.0 Wt. % as Gaseous Olefins 8.85 12.4 8.95 9.6

Wt. % as Inerts 2.18 5.3 2.85 9.94 Wt. % as Liquid Polymers 33.27 52.9 33.2 65.4 Remarks -PACE4. - OF" OIMA-

S e 43 ne

ewj

o 6o4o a /Oo 60 Zo 22

-TM J1 X -D-r A D

I-7

* 0.

I b

-s-- -- -

ibw -6 : Nl ApuCT LOT7 7D RU3 M2R RA )=R f-.A- -O/V6T w.AO \

SDD

-P N.;'M EA?\1' I

RONF-7g 4 RL)V 7A R /

C2H4 RAEMW4A///VO g- G5 POL yW 7a5 7-0 0 as -J------

__ AC TR 7

e- -p

-2 4

Il) -. AE ./ ! - -E %'

S- D D.3 -

I. t

- t - - -- -

44 -

6 ,e1

'TV

e 't.: 4 4 *Al 16

CONCLUSIONS

1. Ethylene gas polymerizes readily under suitable conditions of pressure and temperature without the use of

catalysts.

2. The presence of an amount of naphthalene insufficient

to produce a separate liquid phase increased the amount of

polymerization.

3. The presence of a liquid phase of naphthalene does not

seem to increase the polymerization above that in the case

where there was a gas phase of naphthalene with the ethylene. 4. An increased temperature causes more ethylene disappear-

ance and a larger percentage of heavy liquid polymers.

5. Ethylene disappearance increases oontinuously with time

but gives evidence of approaching a constant value different

from 100%.

6. The polymerization of ethylene starts as a second order

reaction but becomes more complex as higher polymers are

formed.

7. Gaseous products of polymerizations in the presence of naphthalene include olefins other than the ethylene, and hydrocarbon gases which are inert to bromine.

8. The -difficulty of good temperature control when only a

gas phase is present prevents the comparison of reaction 17

rate with that where a liquid phase is also present.

9. Duplication of data is possible as was found by the close checking of runs under identical conditions.

10. Pressure readings during reaction suggest the conclusion that there is an induction period before polymerization begins when ethylene alone is used, and the converse conclusion that with naphthalene present most of the polymerization takes place during the early part of the reaction period.

11. A large iron surface, obtained with steel wool, very decidedly inhibits polymerization of ethylene under the conditions of this investigation.

12. An empirical equation of the second order can be fitted to the data of this investigation at a constant temperature of 7800 F. The integrated form is:

E 2 E1 where E is the per cent ethylene left, 9 is the time of reaction in minutes, and K is a constant equal to

.00064. 18

RECOMMENDATIONS

In making recommendations for further work on this problem the following are suggested:

1. More work should be done on ethylene alone, the results of which could be compared with liquid phase data.

2. Further work with a liquid phase should include short runs at a constant time and varying temperatures so that the variation of reaction rate with temperatures could be observed.

3. Different solvents should be studied.

4. The effect of the filling pressure might be investigated.

5. Any effect of the amount of liquid phase present might be investigated, but it appears to be negligible.

6. Since such striking effects were noticed with an increased iron surface, this should be further investigated.

Improvements in apparatus for this type of work would be advisable for any further investigation. These are as follows:

1. The use of iron cones and iron tubing where there are severe temperature conditions would doubtless reduce trouble from leaks while under operation, but there would be 19

an increased danger of shattering in case of explosion.

2. It would be advisable to use a reactor and method of heating which would afford more uniform temperature conditions inside the reactor. This is particularly necessary in the runs where there is no liquid phase, because the gas does not mix readily, and local overheating or underheating may occur in the system. This could best be done by having a uniform, constant temperature source of heat since direct agitation of the gas is impracticable with a high pressure system. APPENDIX 20

DESCRIPTION OF APPARATUS

For protection against possible explosion the reactor and its accompanying apparatus were mounted behind a sheet steel booth of three sides. It was about five feet high and securely fastened to the floor. The outsides served as operating panels and mountings for control apparatus. High pressure connections were mounted on the inside, and valve handles extended through holes to the outside.

The hookup of the various component parts of the apparatus is indicated on an accompanying flowsheet.

Reactor and Thermocouples The reactor was available and was made from mild steel high-pressure pipe. The flanges and heads were cut from round stock and held by vanadium steel bolts. The heads were pressure-sealed with thin copper gaskets. Details of construction are given on an accompanying scale drawing.

Two thermocouples were used for temperature measurement.

The inside copper-constantan couple was made from high pressure copper tubing with a fine constantan wire in the bore. The couple, which extended into the center of the reactor, was soldered at the tip with "Sil-Fos", and the constantan wire was insulated from the rest of the copper tubing by a Pyrex glass capillary. A chromel-alumel couple was 21

placed on the outside wall of the reactor. It was made by fusing No. 30 chromel and alumel wires in an arc. Both couple leads were insulated outside of the reactor with asbestos tubing. For the cold junction the wires were taken directly to U tubes filled with mercury and placed in a water bath outside the reactor. This made it easy to disconnect the couples when the reactor was taken down without affecting the junction. The temperature of the water bath was recorded with a thermometer. A portable Leeds and

Northrup potentiometer was used in measuring the E.M.F., and contact was to the couples through a double-pole double throw switch. The heater was made by winding 34 feet of 1/8 inch chromel ribbon (0.207 ohms/ft.) over a layer of asbestos paper and a layer of alundum cement. The winding was covered by another layer of alundum cement and wound with asbestos tape. The body of the reactor was insulated with removable pipe lagging. The head with no connections to it was insulated with a cap which slipped on and off and was molded from asbestos paper and magnesia. The other head was wrapped with asbestos paper and tape.

The reactor was swung in a wooden frame with brass bearing surfaces. The axis was a pipe connected at right angles to the center of another pipe running the length of 22

the reactor. The long pipe was fastened into plates

bolted to each head. Shaking was accomplished by fasten-

ing one end of a rod to the reactor and the other end to

the edge of a face plate mounted on a shaft. The shaft was

driven by an electric motor through reduction gears to give

a complete cycle of swing about every five seconds. Connections

The fact that naphthalene was used as the solvent

necessitated the use of heated lines and valves. Both lines

from the reactor were covered with asbestos tubing. A

calculation assuming a reasonable coefficient of heat trans- fer from the tubes indicated a heat input of about four watts per foot. The total length of the main lines was about fifteen feet. These were wound with fifty-five feet of 1/16 chromel ribbon with a resistance of .68 ohms per foot, and covered with asbestos tape. A trap of about 2 c.c. capacity was installed in the high pressure gauge line and was heated with eight feet of the same chromel ribbon, connected in series with the line heaters. Direct current was supplied at 110 volts to the line heaters. All the block valves and both T-blocks were heated with 110 volt-150 watt strip heaters connected two in series to give a 38 watt heat output. These were bound on the blocks with asbestos tape. A glass trap was installed at the vent-off valve to catch the naphthalene 23

before it entered the cold vent-off line leading to the collection bottle. Valves

The valves used were of the standard high pressure block type packed with graphite and asbestos, except the low pressure gauge valve which was a standard Hoke valve.

Gas Collection

For collecting the gas a three gallon bottle filled with ethylene-saturated brine was used. The gas displaced the brine, and the outlet tube was connected by rubber tubing so that it acted as an adjustable barometric leg to assure atmospheric pressure in the bottle. For putting the gas bottle under a head and for refilling, an auxiliary brine bottle was used also.

Analytical Apparatus

The analytical apparatus is described in the discussion of analytical procedure. PHOTOGRAPHS OF APPARATUS LOZ 0W6HEET -FOF APPARPATU Ls

EA Cf- OD -RM-AL LUMEL

OW'1 PREE ,5';,UR

GA 0 E TU-M

O TA /VT A/, ALVE T / E RMFO L UPL

p CC/- VA l- 51IN NA PH THA L E A/Er III 7 C? A P-

GR ADU(A T ECD C VL/DE R D IAGRAM OA

pse3 urs-'F'blE VI\ EWT

JOLN 24

pETAILS OF PROCEDURE

Operating Procedure

Due to the fact that it was found impracticable

to fill the reactor with melted naphthalene under pressure, which would have been necessary if the reactor were not to have been taken apart, one head of the reactor was removed after every run. The used naphthalene was removed, the reactor cleaned out and fresh naphthalene added. The head bolts were then securely tightened so that the copper gas- ket would flow and seal the reactor. Arbitrarily, 100 grams of naphthalene were added since this amount afforded sufficient liquid phase at the operating conditions. The reactor was then remounted, and ethylene was added at a pressure of 190 pounds per square inch by the low pressure gauge, which corresponded to 201 #/sq. in. abeolute pressure.

In this manner conditions of pressure, and therefore concentration, were kept nearly constant at any given temperature throughout the series of runs. Before filling, the reactor was pumped out with a Cenco Hy-Vac pump, filled with ethylene under a slight pressure, and then vented to remove all traces of air in the system. After mounting, the shaking apparatus was started, and the heat applied slowly until the naphthalene was com- pletely melted, as indicated by the inside thermocouple. - -

The temperature was held low until the pressure gauge

indicated that equilibrium had been reached in ths solu-

tion of ethylene into the naphthalene. Heating was then

continued rapidly until the predetermined operating temp-

erature was approached in the reactor. The heat was then

cut down to prevent over-shooting by switching from 220v.

to llOv. and regulating the rheostat. The ammeter in the

line was used so that heat input could be kept constant.

It generally took about 12.5 amperes to hold the tempera-

ture constant. A temperature difference of about 1200 F.

between the inside and outside thermocouples was found nec-

essary to maintain the correct temperature in the reactor.

Direct current was used in all heating to eliminate possible

stray current effects on the thermocouple from alternating

current.

Temperature, pressure, and time readings were

taken at frequent intervals during a run. After the proper

length of heating time, the heat was shut off from the re-

actor, leaving the lines and blocks hot so that they would not clog with naphthalene. The outside lagging on the

reactor was quickly removed and a fan turned on the reactor

to facilitate rapid cooling. Shaking was continued for a while until the temperature had fallen low enough to prevent any localized overheating. The reactor was allowed to cool - 26

until the naphthalene had solidified.

When the reactor was cool, venting off of the

gas was begun by slowly cracking the valve and allowing

the gas to displace the brine in the collection bottle.

The brine was collected in a graduated cylinder until the

pressure in the system had been brought to atmospheric, and was measured to determine the volume of the gas. The

temperatures of the gas in the bottle and in the reactor

were recorded at this time. Knowing the conditions of

pressure and temperature both before and after the run, and

knowing the volume of gas in each case provided data for

accurate calculations after the analysis of the gas was known.

To check calculations on the volume of the system,

several gas determinations were made by filling with ethylene

to a known pressure and temperature and then collecting the

gas at atmospheric conditions both with and without naphtha-

lene present in the reactor. One determination was made by

melting the naphthalene and then allowing it to cool before venting off the ethylene. This was to verify the assumption

that a negligible amount of ethylene remained dissolved in

the solid naphthalene.

Analytical Procedure The analysis of the products was faitly simple

since no attempt was made to identify all of the constituents. 27

The first analysis was the volumetric determination of ethylene and other olefins. This was carried out in a regular Williams gas analysis apparatus. Brine saturated with ethylene was used for manipulating the gas. The absorption bulbs contained 87% H2804, 5% KBr solution saturated with bromine, and a 20% caustic solution respectively.

The 87% acid preferentially absorbs all olefins other than ethylene, although a trace of ethylene is absorbed by it for which correction was made.

A sample of gas slightly less than 100 c.c. was drawn into the burette after thoroughly flushing the lines.

It was passed through the acid until successive burette readings showed a constant difference. This difference was taken as the ethylene absorbed during each pass, and the net difference was the other olefins formed during the polymerization. The bromine would absorb all the olefins, but with the polymers removed the bromine absorption gave the amount of ethylene left in the gas. After several passes through the bromine, the gas was passed through the caustic solution to remove the bromine vapor, and then it was measured in the burette. This was repeated to constant readings. In every case there was some gas left after the ethylene was removed.

It was thought to be a mixture of saturated hydrocarbons and hydrogen, but for convenience was termed the "inerts". .. ~ -

The rest of the analysis consisted of gas

densities and combustions. The densities were measured

gravimetrically by weighing & known amount of the gas. Two bulbs were used, one holding about 250 c.c. and fitted with one stopcock of glass, and the other holding about

100 c.c. and fitted with two stopcocks. For each analysis

the bulb was weighed evacuated and then with air in it,

the difference being the weight of that volume of air. The

bulb was then re-evacuated and filled with the sample and

then reweighed. If the temperatures of each weighing were

the same, no corrections were necessary, and the molecular

weight of the gas sample was determined directly by compar-

ing with the molecular weight of air, 29. A descicator was

used to assure constancy in the weight of the empty bulb.

First the gas as taken from the reactor was weighed, and

then the so-called inerts were removed and weighed separately.

This was accomplished by passing the gas slowly through an

efficient bromine tower followed by a caustic tower. In this

way the olefins were removed before passing into the evacu-

ated bulb. Atmospheric pressure in the bulb was assured by filling under a slight pressure obtained by a few inches

head of brine and then venting off to atmospheric pressure

just before weighing. Therefore, knowing the molecular weight of the ethylene used, and the average molecular weights of 29

the gaseous polymers were calculated from a material balance on the gas.

A few combustion analyses on the inert gases were attempted, but only one or two were successful. A

Shepard combustion apparatus was used for this but due to leaks around some of the stopcocks and inexperience of the experimentors in the technique of the operation of the apparatus, the results were not very conclusive. A measured volume of gas was ignited in a measured excess of oxygen.

The resulting gas was measured to determine decrease on ignition and then passed through caustic to determine carbon dioxide. The remaining gas was supposedly all oxygen, but when absorbed in pyrogallic acid there was found to be a trace of gas left which was considered as nitrogen from air trapped in capillaries. From the resulting data, hydrogen- carbon ratios were calculated using material balances.

These three types of analysis gave sufficient data to make the necessary calculations and conclusions. In addition, however, a semi-qualitative distillation was made on the naphthalene from the reactor containing the non-vola- tile polymers. A variation of an Engler distillation was used, employing a steam jacketed condenser. It was found that no polymers were distilled before the naphthalene, and the naphthalene distilled over quite close to its boiling .1

point (2180 C.). When the last few cubic centimeters

remained in the flask, the boiling point rose rapidly to

an end point of about 2500C. when the flask was dry.

These heavy polymers seemed to be an oil of about the same

quality as a gas oil and varied with the different samples

in quantity rather than in quality. The distilled naphtha-

lene was not pure white, but it was much lighter in color

than before it was distilled. The residue was conversely

a much deeper brown than the original sample. The naphtha-

lene samples all had more or less of an odor of untreated petroleum products. - I

CALIBRATIONS

Thermocoules In determining the calibrations of the thermo-

couples, the authors were fortunate to be able to use a

secondary standard chromel-alumel couple calibrated by

Mr. Walter Ullrich. Both of the couples used in the appar-

atus were put into a thermostatically controlled electric

furnace, and readings made at several points. Unfortunately

the copper-constantan junction was short circuited before

being used and another had to be prepared. This was calibrat-

ed at the boiling points of water and of di-phenyl, and later

at higher temperatures using the secondary standard previously

referred to. In reading the temperatures from the chromel-

alumel couple, the data as given in the chromel booklet of

the Hoskins Company was used. A plot of this was available

through Mr. Ullrich. Data for copper-constantan from the

I.C.T. was plotted, and the curve for our junction was plotted with this as a standard. The curves are shown below.

Gages The low pressure gage was calibrated up to 300 #/sq.in.

the total range, by means of a dead weight gage belonging to

the Mechanical Engineering Department of M.I.T. Points were

taken every 5# for the first 10G# and every 20# thereafter

to 300#. Two such calibrations were made, the second after 32

the gage had been distorted by accidentally putting it under high pressure.

The high pressure gage was calibrated against the low pressure gage in the low ranges. This was accomplished by opening both instruments to the system after they were mounted. For the higher pressures the dead weight gage of the Chemical Engineering Department in the high pressure laboratory was used. Calibration curves for both gages are shown below.

Reactor Volume The volume of the reactor was determined in three ways. It was first calculated from the dimensions of the reactor and the length of tubing used in the connections.

It was then measured by filling the reactor with benzene and then measuring its quantity. Then a gas determination was made by filling the reactor with ethylene under pressure and measuring the amount vented off. By the use of Mu Charts, available through the Department of Chemical Engineering, the reactor volume was calculated. This was done both with 100 grams of naphthalene in the reactor and without any naphthalene. The results of these calculations are shown below. Results of Reactor Calibrations Volume from dimensions; 520 c.c. (Tubing included)

Volume from liquid measure: 512 c.c. (No tubing included) 33

Volume from ethylene measurement: 534 c.c. (Tubing

and low pressure gage line included, no

naphthalene.)

Volume of gas space left in system when 100 grams

of naphthalene are present:

#1. 431 c.c.

#2. 439 c.c.

#3. 432 c.c.

#4. 421 c.c. (Naphthalene melted and then cooled.)

Average value: 431 c.c.

These results were considered as checking well enough since the pressure readings were accurate only to approximately one pound. /''A OF HEa-nNG. C$URVE:. T rrperarvres Refer to InsIde .-apTermocadple...

'APH T~HAL- ENE

7Dg

- - - ** - 7F MECREA4C IN EMv ENA71P-)q RE &sed. on Assumption thatr Rite Do~u e, A6 very ZS5O*

Z'' I9

'p a

406s 6 0 45,5 40 7 5 73o 5 7$5 605

EMVPERATURE DE-6REE CK S- ig W *& Aw T-

RIn Oe ae;

-AREA C /VE VARPE:.=0 WIN x A4M

30 7 0 60 70 80 9

JcE K.T U/ 0

COPPE0 6CO/1 STANAN HEfRqMOCOUPLf CA 4'/BRA TI/ON A

4 99

REC 3

Ibo 200 3650t L..

- -%

INS Uo LwPROSunE GAGA J

-- - -

-9 * -

95 t S 15/ 5 0 z; XAI

r -' -p------4, -

4. * ...... 34

Sample Calculations

1. Reactor Volume:

Temp. in reactor = 690 F. = 5290 R. Temp. of gas collected = 690F. = 5290 R.

Partial press. of water over the brine = 13.5 mm Hg.

Gas collected = 6062 c.c. : 6062 x 746.5/760 = 5960 c.c.

Estimated volume of reactor = 430 c.c. + 90 c.c. of

naphthalene = 520 c.c.

5960 + 430 = 6390 c.c. at 5290 R. & 1 atmos. press. Volume of gas in system at standard conditions:

6390 x 492/529 = 5970 c.c. = .2110 cu.ft. By use of Mu Charts:

Pr = P/Pc = 201/50.9 x 14.7 = .269 V =f NRT/P Tr = T/Tc = 529/508 = 1.04

= .917 V = .917 x .211 x 1543 x 529/359 x 201 x 144 .01523 cu.ft. 431 c.c. 2. Effective Time: (See curves for graphical integration,Figs.7,and 8.

Plot rate of reaction vs. temperature using the assumption that the rate doubles every 250 F., and let the effective

temperature have a rate of 1. Then the area under the

curve equals the effective time at the constant tempera-

ture which was given the rate of 1.) 35

3. Volume of Gas in: (Data from Run #9.)

V = .01523 cu.ft. P = 201 #/sq.in.= 28950 #/sq.ft.

T = 820 F. Tr = 1.0 6 5 Pr = .269 = .91

N = PV" RT = 28950 x .01523/.91 x1543 x 542 = .00058 mols.

Gas in = .00058 x 359 = .2083 cu.ft. Ethylene in = .2083 x .993 = .207 cu.ft. 4. Volume of Gas out: (Data from Run #9.)

Gas vented off 2522 c.c.

Barometer correction = 29.66/29.92 = .992 Partial pressure water over brine 20.5 m.m. Hg

Vol. = (2522 x 739 x 492/760 x 545) + (431 x 492/660) .992=

2514 c.c. at S.C.

= 153.3 cu.in. = .0887 cu.ft. % Ethylene in vented gas = 73.6% Ethylene out = .736 x .0887 = .0654 cu.ft.

5. Ethylene disappearance:

In = .2070 Out = .0654 Disappear. = .2070-.0654 =.1416 c.c 100 x .1416/.2070 = 68.4% disappearance

6. Molecular Weights: Bulb evacuated = 73.3589 gms. " + air = 73.6406 " " + gas = 73.6816 1

Molecular weight of gas = (73.6816 - 73.3589) x 29/(73.6406 - 73.3589) = 33.4 Molecular Wt. of Gaseous Olefins: M.W.Gas = 33.4 M.W.Ethylene=28: %=73.6 M.W.Inert Gas=32.95+A17.06 I Gas.Olefins = 9.34 33.4 - (.736 x 28 + .1706 x 32.95) = 7.22 36

7.33/.0934 = 76.9 = Mol.Wt. of Gaseous Olefins 7. Weight Percentages of the Products: Wt. % in Liquid Polymers: In = .2070 x 28 = 5.80 = Lbs. x 359

Out = .0887 x 33.45 = 2.96 = Lbs. x 359

(5.80 - 2.96)/5.80 = .4879 = 48.79% 8. Hydrogen: Carbon Ratio:

Gas sample = 17.90 c.c.

Oxygen used = 55.45 c.c.

CO2 formed = 28.00 c.c.

28.00/17.90 = 1.56 atoms of carbon/molecule of gas

55.45 - 28.00 = 27.45 c.c. of oxygen to water

27.45 x 4/17.90 = 6.14 atoms of hydrogen/molecule of gas

6.14/1.56 = 3.94:1 4 e4 _- Ills. 00I h Ij lIQ a * ,

4, -4)4, ~19

)) 0) 0 M ODt ,-'r C6

e 4 t, o P, o%*4 aC *e e al

% CD ie \4

N Lu

5r 5 * 0) 03 t-

t C-) Oa

-- C a)

* sS -5 4 , e o ~ 0o4 oo eo o 00

-4 1,, c> L) c o e

03 re c u

co Q)o

* *44 c I N ne' t

* o aa *o

cco S 0 INN

4 5

S* o C G ,

- * -' N4 4

* SBG

a -C

* ,

5 *4* 5**3 C 43 *

) C4

1 S1

*40 A Po 1, PE S UR EL EPHA - E PATA

'11

SOO -- -

-q - -. *- . -

A - -£ 37

LITERATURE CITED

1. Elkington, H.D.; British Patent 331,186, June 1930.

2. Ramage, A.S.; U.S.Patent 1,527,079, February 1925.

3. Egloff and Schaad; J. Inst. Petroleum Tech., 19; 800-811

(1931) 4. Ellis, C.; "The Chemistry of Petroleum Derivatives" The Chemical Catalogue Co., New York 1933.

5. Francis and Kleinschmidt; Oil Gas J. 28, 118, (1929). 6. Frolich, Tauch, Hogan, and Peer; Ind. Eng. Chem. 23; 548-

550, (1931).

7. Hague and Wheeler; J. C. S. 390, (1929). 8. Ipatiev; Ber., 44; 2978, (1911).

9. Lewis, W.K.; Trans. Am. Inst. Mining-Met. Engrs. 17; 1-22, (1934).

10. Nash, Stanley, and Bowen; J. Inst. Petroleum Tech. 16;830-

857, (1930).

11. Pease, R.N.; J. Am. Chem. Soc. 52, 1158-1169, (1930).

12. Sachanen and Tilicheyev; Ber. 62, 658, (1929). 13. Sachanen and Tilicheyev; "Chemistry and Technology of

Cracking", The Chemical Catalogue Co., New York 1932.

14. Tropsch and Philipovitch; Brennstaff Chemie,4;147-149,(1923).

15. International Critical Tables.